Object classification and identification from raw data

Abstract

Raw data is accessed from a storage device. A sample survey technique is used on the raw data to select a sample data. A data science technique is used on the sample data to determine a sample data category. The raw data is classified at least in part by considering the sample data category. A tier of storage is identified for the raw data on the storage device based on the classification.

What is claimed is: 
     
       1. A method, comprising:
 unilaterally accessing a raw data from a storage device; 
 using a sample survey technique on the raw data to select a sample data; 
 using the sample survey technique on an updated raw data to select an updated sample data; 
 using a dynamic data science technique comprising a state change analysis on the sample data and the updated sample data to determine a sample data category; 
 using a processor to classify the raw data at least in part by considering the sample data category; 
 iterating to adjust the sample survey technique based at least in part on the classification of a nearby sample data, wherein adjusting the sample survey technique includes increasing confidence and decreasing sample density of future sample surveys for a same LUN; 
 providing the classification to a tiering software associated with the storage device; and 
 adjusting classification on future sample data categories for the same LUN. 
 
     
     
       2. The method of  claim 1 , wherein the dynamic data science technique is one or more of the following: a Laplace transform, a Fourier transform, and a Markov chain. 
     
     
       3. The method of  claim 1 , wherein the storage device is one or more of the following: a storage array, an element of the storage array, and a LUN. 
     
     
       4. The method of  claim 1 , wherein the sample survey technique uses simple random sampling. 
     
     
       5. The method of  claim 1 , wherein the sample survey technique uses data analytic software. 
     
     
       6. The method of  claim 1 , wherein the sample data category includes one or more of the following: MP3s, MP4As, ID3, mhod, MP4, WMV, SQL files, database files, music files, video files, photo files, database files, database logs, logs, classification objects, and objects used in other classification projects. 
     
     
       7. The method of  claim 1 , wherein the tier of storage includes one or more of the following: a Tier-0 drive, PCI Flash drive, SSD drive, Tier-1 drive, FC drive, 15 k rpm Serial-Attached SCSI drive, 10 k rpm Serial-Attached SCSI drive, Tier-2 drives, Nearline Serial-Attached SCSI drive, SATA drive, Tier-3 drives, Tape drive, and a cloud storage. 
     
     
       8. The method of  claim 1 , wherein the data science technique uses data analytic software. 
     
     
       9. The method of  claim 1 , wherein the data science technique includes micro-text mining. 
     
     
       10. The method of  claim 1 , wherein the data science technique includes clustering using statistical fingerprinting. 
     
     
       11. The method of  claim 10 , wherein clustering using statistical fingerprinting comprises assigning an observation to the nearest statistically close cluster. 
     
     
       12. The method of  claim 10 , wherein statistical fingerprinting includes using one or more of the following: a Hamming weight, an arithmetic mean, a Shannon entropy, and a Kolmogrov-Smirnov p-value. 
     
     
       13. The method of  claim 1 , wherein adjusting the sample survey technique includes using a range iteration routine. 
     
     
       14. The method of  claim 1 , further comprising decrypting the sample data before using the data science technique. 
     
     
       15. The method of  claim 1 , further comprising adjusting the data science technique based at least in part on using a machine learning algorithm over a plurality of previous classifications. 
     
     
       16. A system, comprising:
 a processor configured to: 
 unilaterally access a raw data from a storage device; 
 use a sample survey technique on the raw data to select a sample data; 
 use the sample survey technique on an updated raw data to select an updated sample data; 
 use a dynamic data science technique comprising a state change analysis on the sample data and the updated sample data to determine a sample data category; 
 classify the raw data at least in part by considering the sample data category; 
 iterate to adjust the sample survey technique based at least in part on the classification of a nearby sample data, wherein adjusting the sample survey technique includes increasing confidence and decreasing sample density of future sample surveys for a same LUN; 
 provide the classification to a tiering software associated with the storage device; and 
 adjust classification on future sample data categories for the same LUN; and 
 a memory coupled to the processor and configured to provide the processor with instructions. 
 
     
     
       17. A computer program product, the computer program product being embodied in a non-transitory machine-readable storage medium and comprising computer instructions for:
 unilaterally accessing a raw data from a storage device; 
 using a sample survey technique on the raw data to select a sample data; 
 using the sample survey technique on an updated raw data to select an updated sample data; 
 using a dynamic data science technique comprising a state change analysis on the sample data and the updated sample data to determine a sample data category; 
 classifying the raw data at least in part by considering the sample data category; 
 iterating to adjust the sample survey technique based at least in part on the classification of a nearby sample data, wherein adjusting the sample survey technique includes increasing confidence and decreasing sample density of future sample surveys for a same LUN; 
 providing the classification to a tiering software associated with the storage device; and 
 adjusting classification on future sample data categories for the same LUN.